On Mon, Jul 11, 2016 at 04:06:25PM +0200, Takashi Iwai wrote: > On Mon, 11 Jul 2016 16:07:32 +0200, > Vinod Koul wrote: > > > > > > diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c > > > > index 9a0d1445ca5c..b4ebdde59398 100644 > > > > --- a/sound/pci/hda/hda_intel.c > > > > +++ b/sound/pci/hda/hda_intel.c > > > > @@ -1649,6 +1649,7 @@ static int azx_first_init(struct azx *chip) > > > > return -ENXIO; > > > > } > > > > > > > > + azx_parse_capabilities(chip); > > > > > > Is it really safe to call the function no matter which chip? > > > > I think so, since we read base HDA registers for capablity. For older hw the > > next capablity will not be there. > > > > But yes I will verify this on older machines. > > I checked the spec 1.0 and 1.0a, and there is no definition for the > register offset 0x14. So, it's likely unsafe to use it > unconditionally. Ah okay, let me check. Worst case we cna use SKL_PLUS for this as it is defined for SKL and beyond. -- ~Vinod _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxx http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
